Generator Charging

[ajdelange]

Yes, I think that's pretty clear. I think you could even use a "cheap" generator as any rotary machine produces a pretty good sinewave (10% or so distortion). Now I suppose you could run a cheap generator at any speed to turn an alternator, rectify that and send the DC to a modified sine wave inverter but I'm not aware of anyone doing that. Many "inverter" generators are sold these days and it's implied that they are "pure" sine inverters but not many of them actually give a THD spec.

Advertisement

 

[Bill906]

Sorry, I thought I addressed the VCO part, but I did not.

Yes a Voltage Controlled Oscillator (VCO) can lock onto a very crappy waveform. But an active converter that needs to synchronize with the AC line is not the same thing. The active converter needs to turn the transistor on and off at the correct times with respect to the instantaneous voltage of the AC line. With a sine wave there is a relatively gradual change in voltage over time compared to a square wave which is abrupt. One microsecond the square wave is full negative, the next microsecond it’s full Positive. With a modified sine wave one microsecond it’s at zero volts, the next microsecond its at full positive. Not sure the gate firing algorithms were designed to be able to actively adjust to such unpredictability. They are not just turning on and off to rectify the AC into DC, they need to turn on and off to mitigate harmonics, correct for power factor, boost the DC voltage level above peak of sinewave and although the current Tesla chargers don’t do this, an active converter also can send power from the batteries into the active AC line. I just don’t see the charger being able to do this with a line voltage waveform other than a decent sinewave.

I‘m not sure your test of using a pretty good, but not perfect sinewave power generator tells us if a Tesla can charge from square or modified sine wave power generator.

Just curious, what are your thoughts as to why Tesla cautions against using private generators?

I have worked with Variable Frequency Drives in the industrial world that have active converters and have seen them fault out on “Line Sync loss“ if a significant power distrurbance occurs. If the distubance isn’t something that can be fixed we program the drive to continue running but stop the active converter until it can resync with the line. Durning that time the converter is only a passive converter (flyback diodes in parallel with the transistors). Next time I’m working with a development engineer for that product I‘m going to ask him what it would do if the AC line was a square wave.

 

[ajdelange]

Sorry, I thought I addressed the VCO part, but I did not.

You did address it here...

I believe the phase lock loop (PLL) algorithm would need to latch on to a true sinewave voltage waveform to time that gating correctly.

Now you say..

Yes a Voltage Controlled Oscillator (VCO) can lock onto a very crappy waveform.

Then you say..

But an active converter that needs to synchronize ... Tesla chargers don’t do this, an active converter also can send power from the batteries into the active AC line. I just don’t see the charger being able to do this with a line voltage waveform other than a decent sinewave.

So why are you arguing about the limitations of a system which, and you argue this passionately, does not have an active inverter? Do you see why I find it hard to follow you?

I‘m not sure your test of using a pretty good, but not perfect sinewave power generator tells us if a Tesla can charge from square or modified sine wave power generator.

Well no, it doesn't but it does tell us that the Tesla power supply, as is required of all power supplies made today, can deal with and improve upon input waveforms with some amount of THD. It doesn't tell us what the limit of tolerable THD is though.

It appears that the current Tesla architecture currently uses passive rectification of the incoming waveform dumping the rectified pulses into a capacitor. Dumping the output of the first bridge into a 1000 ufd capacitor and discharging that at an 11.5 kW rate would give voltage across the capacitor that looked something like this:

First a comment on the huge ripple. The fact of the matter is that if you are going to be discharging a capacitor at a rate corresponding to 11.5 kW then it is going to discharge pretty fast unless the capacitor is huge. 1000 ufd capacitors rated at 500V are certainly available but they aren't cheap. The fact is that 10,000 W is 10,000 joules/sec and a 1000 ufd cap at 400V only holds 90. The glaring thing about the voltage curved is that from the POV of ripple the modified sine wave produces a much better result. And while we have demonstrated that phase lock with the primary waveform is clearly possible it should be clear that sync to the the 120 Hz line in the rectified waveform would be even simpler. So there is no problem with sync and the rectified modified sinewave is purer DC it seems that there would be absolutely no problem with charging a Telsa from a modified sinewave inverter. But no one has asked about that. The question is as to whether it can be done from a generator. It clearly can be done from a pure sine wave invertrer equipped generator and it clearly can be done from a conventional generator. It looks as if it can be done from a generator equipped with a modified sine inverter too but I have never seen or heard of such a beast.

Just curious, what are your thoughts as to why Tesla cautions against using private generators?

Obviously came from the lawyers. Some Harry Homeowner fiddling around with a generator is going to hurt himself or damage something. It's for their protection.

I have worked with Variable Frequency Drives in the industrial world that have active converters and have seen them fault out on “Line Sync loss“ if a significant power distrurbance occurs. If the distubance isn’t something that can be fixed we program the drive to continue running but stop the active converter until it can resync with the line. Durning that time the converter is only a passive converter (flyback diodes in parallel with the transistors). Next time I’m working with a development engineer for that product I‘m going to ask him what it would do if the AC line was a square wave.

Seems reasonable but as the current Tesla circuit isn't active its moot.

 

[Geo]

I travel to South Central Ohio occasionally, and chargers are few and far between. 500+ miles should keep me moving, but still not comfortable with the idea.
Having a bed opens up some possibilities.

My thoughts are to throw a Pure Sine Wave gas generator, with 220V, in the Vault as an emergency backup. I hear Teslas are picky about sine wave cleanliness, so a Pure wave would probably be best. It would take a couple hours at 220, but it's better than a tow.

Obviously using gas as a backup will upset some "purist", but I don't care.

I could use the generator around the house, and camping, so it wouldn't be solely for travel...

Anything more I should be looking at for this plan?

Hey Ryan,

I'm with you on your generator solution and why its necessary.
I think someone got way hung up in the weeds with your mention of electricity line cleanliness, and missed the whole point.
( I think they were trying to dissuade you from using a generator, but lets be kind and suggest that they were giving you details to build a generator instead of buy one )

The simple fact is, as you have already highlighted, and implicit in your question, YES, there are many Generators available that can do the job, the internet in general and YouTube particularly has many examples of what will work.

Unfortunately some people tried hard to deflect from that fact, and didn't give you any practical info, (and me either ;-).

I'm not aware of any legal reason that Tesla put forward to stop you using a generator.
I don't believe its in their power (no pun intended ) if they wanted to !

Please post up if you zero in on a model

 

[Aces-Truck]

My thought on the disclaimer from Tesla is twofold:
1) As has been mentioned, it limits their liability.
2) the last thing Elon Musk wants is folks with Range Anxiety to put a generator on a trailer, and turn their EV into a fossil fuel burning car.

 

[Geo]

My thought on the disclaimer from Tesla is twofold:
1) As has been mentioned, it limits their liability.
2) the last thing Elon Musk wants is folks with Range Anxiety to put a generator on a trailer, and turn their EV into a fossil fuel burning car.

Hi Kevin, can you please provide the Tesla disclaimer you're referring to ?

The Tesla staff here have previously advised they don't have such a disclaimer.

As it has been put to them that a premises (or remote campsite) may be powered by a generator, and as such any other device that was powered by the premises previously will continue to work as usual under the generator (provided its sized sufficiently. If not, no harm or danger exists, the Tesla simply cuts it self off from the power source provided. )

 

[ajdelange]

Hi Kevin, can you please provide the Tesla disclaimer you're referring to ?

It's mentioned in No. 27:

I did find that the caveat against "private generators" still exists -not in the manual for the car where I had been looking but in the manual for the Universal Mobile Connector (as it is called) and fot the Gen 2 and 3 HPWC. "CAUTION: Do not use private power generators as a power source for charging".

The wording is certainly peculiar. If I have a Honda generator I shouldn't, per Tesla's caveat, use it. But if FEMA lends me such a generator in the aftermath of some hurricane or tornado then it is OK to use it?

As it has been put to them that a premises (or remote campsite) may be powered by a generator, and as such any other device that was powered by the premises previously will continue to work as usual under the generator (provided its sized sufficiently. If not, no harm or danger exists, the Tesla simply cuts it self off from the power source provided. )

I can imagine the arguments in court "But m'lud, the generator is owned by the campground which is operated by a publicly traded company."

I had a small Generac unit that had pretty poor voltage regulation (you could see the lights flicker). The Tesla charger looks for fluctuating voltage and reduces charging current if it sees it on the premise that the fluctuations are caused by poor pin contact on one of the plugs. I never charged the Tesla with it but such a unit might present a problem. A properly regulated conventional generator should present no problem.

The remaining question in my mind is as to whether there is a limit on THD that is acceptable to the car. The portions of the manual that discuss these features do not mention one. Suppose the camp in question has no generator but rather operated on solar panels. Pure sine inverters are expensive. For most of the loads modified sine is fine. Should we not charge our cars at such a camp ground? There is no warning against doing that in any of the manuals and, as the discussions here that go deep into the weeds have shown, no apparent technical reason why one should be concerned.

 

[Geo]

It's mentioned in No. 27:

The wording is certainly peculiar. If I have a Honda generator I shouldn't, per Tesla's caveat, use it. But if FEMA lends me such a generator in the aftermath of some hurricane or tornado then it is OK to use it?

I can imagine the arguments in court . . . . .


Should we not charge our cars at such a camp ground? There is no warning against doing that in any of the manuals and, as the discussions here that go deep into the weeds have shown, no apparent technical reason why one should be concerned.

Apart from your spurious theatrics about being in court, as there is zero reason to be in court,
The Tesla will either accept charge from an adequate generator or it will decline charge
(without any damage being caused to the car or the generator)

Thank you for last post and finally answering the simple question, can you charge your Tesla by a generator.
The answer is yes. (however, that was already a given in the original starting post !!)

And with all your technical knowledge, its amazing you didn't just give an example of a unit that will work, (and you still haven't) before going on your journey exploring things needing a spectrum analyser and oscilloscope. Still, I feel a little deprived you didn't detour into the right atmospheric conditions to charge the car up using a kite in a thunder storm !
As one of your previous answers somehow managed to include the river Nile !

 

[ajdelange]

I did mention that I was able to charge it from an RCL48 Kohler generator (though I didn't mention the generator's make or model number). That's a whole house backup generator and probably not of much interest here. I also mentioned that I had charged from a Jackery 1 kWh battery pack but that's got a sine wave inverter so probably not of much interest either.

Most of what I know about the Australian legal system I have learned from Rake but here in the USA everyone sues everyone for anything. If you have been charging from a generator (a privately owned one, that is) you have given Tesla a bit of ammunition in any legal action that they would not have had you not been. That was the point of the remark about court.

 

[Bill906]

The first converter (or bridge) in the Tesla charger uses transistors. That makes that part of the charger an active converter. The reason that converter is an active converter is most likely for harmonic mitigation. Because of this I find it unlikely the Tesla charger will work with voltage waveforms other than ones that closely resemble sine waves. If you use a cheap generator that creates a squarewave or modified square wave I doubt the Tesla charger will work.

To answer the initial question of this thread, Tesla cautions against using a private generator. From sources online some have had good results. The main issue seems to be that the ground and the neutral need to be bonded. Once that is done most had little issue. I would also recommend using a “Pure“ or “True” sinewave generator. I doubt a Tesla charger will accept the squarewave produced by cheaper inverter generators. If using a non-inverter generator I believe you will have problems unless it is a very good one, or sized much larger than needed for charging the vehicle.

 

[Newton]

The concept of tesla sueing me over how I use my car is a reason I wouldn't get 1.
Also
The fact that they can change and make my car useles via mandated updates is just disgusting. The whole issue of them not allowing fast charging on some salvage title cars is one of the main reasons I might not buy the CT

 

[ajdelange]

To answer the initial question of this thread, Tesla cautions against using a private generator. From sources online some have had good results. The main issue seems to be that the ground and the neutral need to be bonded.

Rather than make people with no interest wade through my explanations as to why the claims in the quoted post are baseless I thought I would start with something that would be pertinent to all who contemplate charging with a generator.

Charging a Tesla from a portable generator presents a hazard because the vehicle is not properly grounded. That is why Tesla has a Caution against the practice. They aren't willing (nor should they be) to give us credit for being able to figure out how to ground it properly. In the many videos I have seen on charging the cars from generators I have not seen one that even mentions this.

Let's start with how it works in your residence. There is a transformer on the pole or in a vault that produces biphase 120 i.e. two hots and a neutral. The neutral is grounded at the transformer and it is also grounded at the service entrance (and nowhere else). That ground is dervived by driving one or more rods into the soil and the ground (bare or green) wire is connected to them (and the neutral where the power comes in. The Tesla UMC plugged into your outlet or any of the HPWC models checks the voltage between L1 and the ground wire. If the ground wire is not connected it won't measure the requisite 120V and won't allow power flow to the vehicle. This is it's way of checking that the vehicle frame, connected to the ground wire, is at ground potential.

To fool the UMC into working it is only necessary to connect the neutral and ground terminals together at the generator. The UMC then measures 120V between L1 and the ground wire, thinks it's connected to a properly wired residence and closes the contactor.

Now suppose the car is being charged from a generator in an area of thunderstorms and that is gets charged to 10,000V realtive to ground by leaders or whatever, If the tires are dry they will make pretty good insulators and the charge may remain on the vehicle for a time. The generator will float up with the car so it is also at 10,00V with respect to ground. You come along and touch either the generator or the car and you get the 10,000 volts.

What's the solution to this problem? Do what they did in your house. Prevent the power system (generator) from floating. Ground it! There is a terminal on it for that purpose. Drive a ground rod and connect it to that or connect it to another ground such as a water pipe, the grounding system of a house... Now when St. Elmo comes around his charge will leak off the car and generator through the ground wire and neither can float above ground. You are safe.

Thus charging advice is:
1)Use whatever generator you have available.
2)Bond the generator's neutral to the generator's frame (grounding terminal)
3)Ground (earth) the grounding terminal.
4)If you ignore 3), don't do it when there is a threat of a thunderstorm.

So why, just based on simple logic, can you use any generator you want? A couple of questions and answers.

Is there potential danger charging with a generator? Answer: Yes
Is there a Caution about doing this in the manuals?A Answer: Yes

If the voltage supplying the UMC is not stable can that interrupt charging: Answer: Yes
Are there warnings about this in the manual? Answer: Yes

Does the manual have warnings about high THD? Answer: No
Would it seem, therefore, that high THD is a problem? Answer: No

A pure sine inverter generator will produce THD of < 3%
The crummiest generator out there will produce THD of 10 - 20%
In general non - inverter generators produce THD of 9 - 10%
Good generators produce THD of 5 - 6%

Now it stands to reason that the cheapest generators out the will not only have the highest THD but probably also be the least stable with the common reason being unsophisticated field control for voltage regulation. Thus we might modify 1) to

1)Use a decent quality generator.

It doesn't have to be a true sine inverter generator which produces wavefoms cleaner than the utility's. Tesla chargers work very well indeed with utility quality power (< 7%). Why would we need to give them power better than the utility's?

The remainder of this post is more technical detail showing why Bill906's assumptions are probably wrong.

The first converter (or bridge) in the Tesla charger uses transistors.

Probably not. That rumor was started when someone looked at the packaging and saw the extra pins. That does not make the things in the package transistors. And even if they were there is no reason why they should be gated other than on other than as diodes. The rumor that the front end bridge devices were transistors led to the conclusion that the device was a two way device. Were that so these devices would have to be transistors. As the charger is not presently two way there is no reason for those devices to do anything other than rectify.

That makes that part of the charger an active converter. The reason that converter is an active converter is most likely for harmonic mitigation. Because of this I find it unlikely the Tesla charger will work with voltage waveforms other than ones that closely resemble sine waves.

Harmonic mitigation is done elsewhere. Several sources which include schematics and reverse engineering attempts make it pretty certain what the architecture of the charger is. It is apparently a passive bridge (diodes or transistors configured as diodes) rectifier followed by a second bridge inverter operating at something like 50 kHz. The inverter output goes to a transformer where the AC voltage is stepped up and the stepped up AC is then rectified and sent to the battery. Given this the sensible place to do "power factor correction" is in the gating to the inverter stage. The desired current draw waveform is stored in the machine and it is generated locked to the incoming voltage. It is compared to the current being drawn by this bridge and an error signal developed. That error is fed to a PWM gating waveform generator. If the charger is drawing too much current at a particular time in the cycle the pulse width is narrowed. If it is not drawing enough it is widened. Note that one can also control the magnitude of the current drawn by controlling the scaling of the reference waveform.

The output of the modulator (bridge) is a series of bipolar (AC) pulses of varying widths that is passed to a transformer where the voltage is stepped up to a level suitable for the battery. That pulse train is rectified (in a third bridge) to provide DC for the battery.

I need to make it clear that this is probably the way the circuit works given what we know about its architecture and the way many a switching power supply works but I don't have enough detail to say that this IS what it does. It's just probably what it does. The inverter bridge has to be gated (it has to be active) to do its function. Given that, it is the logical place to do "power factor" correction and, given that why would they want to complicate things by gating the incoming bridge too?

Because of this I find it unlikely the Tesla charger will work with voltage waveforms other than ones that closely resemble sine waves.

The front end probably is not active but even were it so what would be the problem with highly distorted inputs? Your argument is jejune.

If you use a cheap generator that creates a squarewave or modified square wave I doubt the Tesla charger will work.

There are no generators that create square wave or modified square wave and even if there were they would probably work. For perspective here's a picture of three voltage waveforms each of which has the same rms voltage (0.707 volt). The heavy red curve is a pure sine wave - 0% THD. The blue curve is a distorted waveform with 20% THD and the stepping curve is modified sinewave with THD 29%. Why do we think a rectifier, active or passive, would have any trouble with any of these?

There are limitations, of course, as to how much waveform correction can be done by this method. Clearly you cannot draw current when the supply voltage is 0 as it is for about 1/3 of the modified sin inverter's cycle. Keep in mind that an OEM is not obligated to deliver better than a 7% THD waveform to a generator. He is obligated to draw a current waveform that has less than 7% THD when supplied a voltage waveform with THD < 7% which is the worst a utility is allowed to supply 95% of the time. This is all for the protection of the utility, btw. Harmonic currents are often reverse sequence and the big killer is that triplen harmonics add up in the 0 sequence (on the neutral) to the point that it now has to be sized as large as the phase conductors.

I would also recommend using a “Pure“ or “True” sinewave generator.

Not necessary as proven above. OTOH the pure sine generators are really big right now. It's a great marketing tool. The generators that I'd consider portable and even some of the heftier "contractor" numbers all seem to be pure sine inverter these days.

I doubt a Tesla charger will accept the squarewave produced by cheaper inverter generators.

Can you point me to one of these? I can't find any.

 

[ajdelange]

The concept of tesla sueing me over how I use my car is a reason I wouldn't get 1.

That's not the danger. The danger is that if something happens to your car while you are charging it with a generator or even at some other time given that Tesla finds out that you have been charging it from a generator (nosy neighbor spills the beans) then they could deny a warranty claim or liability claim based on the fact that you did something the manual specifically cautioned you not to do.

 

[ajdelange]

I doubt a Tesla charger will accept the squarewave produced by cheaper inverter generators.

Would you also think that other manufacturers' switching power supplies would not work on a modified sinewave or is it just Tesla's?

 

[ajdelange]

Charging a Tesla from a portable generator presents a hazard because the vehicle is not properly grounded. That is why Tesla has a Caution against the practice. They aren't willing (nor should they be) to give us credit for being able to figure out how to ground it properly. In the many videos I have seen on charging the cars from generators I have not seen one that even mentions this.

Since I wrote this I did come across one. While they didn't actually tell you how to do it properly they did wave a grounding rod about for a few seconds.